The aim of this research is the development of improved statistical methodology for the analysis of medical studies. The extension and modification of existing methodology will be studied in the context of the analysis of actual data sets. Initial areas of investigation will include the problems of competing risks, the estimation of cure rate and the analysis of case-control studies. A primary source for data sets will be the clinical and laboratory activities within the Fred Hutchinson Cancer Research Center.